How to Make Law School Flashcards

Card the things that have a fixed, checkable answer. That means element lists (the four elements of adverse possession in your jurisdiction), definitions (hearsay under Fed. R. Evid. 801(c) is an out-of-court statement offered to prove the truth of the matter asserted), standards (Rule 12(b)(6) tests facial plausibility; Rule 56(a) requires no genuine dispute of material fact), burdens and who carries them, exceptions and their triggers, and multi-part tests (the two prongs of Strickland v. Washington, 466 U.S. 668 (1984): deficient performance and prejudice).

Do not card things that have no single answer. Policy arguments, competing rationales, the reason Erie R.R. v. Tompkins, 304 U.S. 64 (1938), matters to federalism, how a professor's hypothetical should come out — none of that fits a front-and-back format. You will either write a mushy answer you can't grade yourself on, or you will memorize a sentence that collapses the moment the exam changes a fact. Put that material in your outline and work through it with hypotheticals instead.

A good filter: if you can imagine a bar-style multiple-choice question or a one-line blank on the answer, it belongs on a card. If the honest answer starts with "it depends on how you frame it," it doesn't.

Keep cards atomic — one fact, one card. The single most common mistake is the dump card: front says "Personal Jurisdiction," back is a half-page. You cannot grade yourself on that, so you look at the back, think "yeah, I knew most of that," and learn nothing. Break it up.

Turn a paragraph in your outline into a small family of cards. From minimum contacts under International Shoe Co. v. Washington, 326 U.S. 310 (1945), you might get: "What is the two-part test for specific personal jurisdiction?" / "Minimum contacts with the forum plus fair play and substantial justice." Then: "What must the contacts be purposeful toward?" Then: "Name three of the fairness factors." Each has a discrete answer you either produce or you don't.

Use cloze deletion for element lists and statutory language. Instead of a front/back card, write the full rule and hide one piece at a time: "A promise which the promisor should reasonably expect to induce action or forbearance ... and which does induce such action or forbearance is binding if [injustice can be avoided only by enforcement of the promise]" (Restatement (Second) of Contracts § 90). Cloze forces you to reproduce the operative language rather than paraphrase it into something exam-useless.

Write the front as an actual question, not a topic label. "Excited utterance elements?" is weaker than "What three things must be shown for an excited utterance under Fed. R. Evid. 803(2)?" The second one has a countable answer, so you know when you missed one.

Two smaller mechanics that pay off: put the jurisdiction or the professor's variation on the card when your class departs from the majority rule, and add a short cite line on the back (case name plus reporter, or the rule number) so the card doubles as a pointer back into your outline when something looks wrong.

Make far fewer case cards than you think. In most doctrinal classes you are tested on rules, not on your ability to recite Pennoyer v. Neff, 95 U.S. 714 (1877). Card a case when (1) your professor treats the case name as shorthand and expects you to use it, (2) the case is the rule — there is no clean statutory statement, as with Palsgraf v. Long Island R.R. Co., 248 N.Y. 339 (1928), on foreseeability and duty — or (3) the case is a landmark you would be embarrassed to miss on an exam or in a clinical interview.

When you do card a case, three fields is enough: the name, the rule or holding in one sentence, and the one fact that triggered it. Skip procedural history, skip the dissent unless your professor spent a class on it, skip the full citation unless you're expected to cite it. A card that reads "Miranda v. Arizona (1966) — custodial interrogation requires warnings; trigger: questioning while in custody" is more useful than a card that reproduces your case brief.

In upper-level classes and in seminars where the professor tests doctrinal evolution, this flips: you may need cards for a line of cases and what each one changed. Ask what the exam looks like before you decide.

Anki if you are making several hundred cards you want to retain past the semester, especially with an eye toward the bar. It is free on desktop and Android, paid on iOS, it handles cloze deletion natively, and its scheduler spaces reviews so you see weak cards often and strong cards rarely. That spacing is the entire point — reviewing everything every day is what makes people quit.

Quizlet if you want low friction and shared decks, or if you're carding a single class for a single exam. It's faster to set up and easier on a phone, and the scheduling is less sophisticated, which matters less over fourteen weeks than over three years.

Paper is genuinely fine for one course, and the act of handwriting the card is itself encoding. The failure mode is volume: a 400-card paper deck for Civil Procedure becomes unmanageable to shuffle and impossible to sort by difficulty.

Whatever you pick, do not download a stranger's deck for a doctrinal class and treat it as your deck. It won't match your professor's rule statements, it won't reflect your jurisdiction's variations, and you'll skip the synthesis work that makes cards worth having. Commercial bar decks are a different story — for the MBE, the law is national and standardized, so a professionally built deck is a reasonable starting point.

Use AI to reformat material you already understand, not to generate law you haven't learned. The safe workflow: paste a section of your own outline and ask for atomic question-and-answer pairs or cloze deletions in a format your app can import (Anki takes plain text with a consistent separator). The model is doing clerical work — splitting a paragraph into six discrete prompts — which is exactly the kind of task it does reliably.

The unsafe workflow is asking a model to "give me flashcards on the hearsay exceptions" from nothing. Language models produce plausible-sounding rule statements that are subtly wrong, blend majority and minority rules, cite provisions that don't exist, and occasionally invent case names outright. A wrong card is worse than no card, because spaced repetition will drill the error into you until it feels true.

So verify before a card enters the deck. Check every rule number against the actual Federal Rules text, every case name against your casebook or a real database, and every rule statement against your professor's language. If a card cites something you don't recognize from class, delete it rather than trusting it.

One more thing AI does well: generating practice questions from your cards. Ask it to write short hypotheticals that turn on a rule you've carded, then answer them yourself. That closes the gap between knowing a rule and applying it — which is the gap flashcards leave open.

Cards come after synthesis, not before. Read, brief lightly, attend class, and add to your outline weekly. Then, from that week's outline additions, cut ten to twenty-five cards. Making cards from a casebook you haven't outlined produces disorganized cards and skips the step that actually builds understanding.

Review daily in short blocks — fifteen to twenty minutes, before class or on the walk over. The value of spacing comes from frequency, not duration. A ninety-minute Sunday cram undoes the scheduling.

Budget the last three to four weeks before exams for practice, not card-making. By then your deck should be closed except for corrections. If you are still writing cards during reading period, you are memorizing when you should be applying.

Expect the deck to shrink. Delete cards you've answered correctly five times in a row and cards that turned out not to matter to your professor. A 250-card deck you actually run is worth more than an 800-card deck you abandon in week nine.

They won't teach issue spotting. An exam fact pattern doesn't tell you which rule applies; it buries three triggers in a paragraph about a bar fight. That skill comes from doing old exams under time and comparing your answer to a model.

They won't teach application or structure. Knowing the four elements of a claim is not the same as arguing both sides of the contested element on these facts. Flashcards get you to the point where you don't waste exam minutes retrieving the rule — the rest of the points come from what you do with it.

And they won't help much in a class graded on a paper, an oral argument, or a heavily policy-driven exam. Before you build a deck, find out what the assessment actually is. In a class where the professor hands out a closed-universe problem in week two, your time goes to the problem.